How to explore : Hovering over any point on the map or histogram will link coffee shop locations and sentiment scores. Sentiment scores are computed using text analytics techniques on all text reviews for each coffee shop. Sentiment scores above 0 indicate positive reviews while scores below 0 indicate negative reviews.
Quick takes : The two Joe Coffee outlets have achieved positive reviews, likely due to good locations and good quality coffee, which is great news. But more interestingly, notice that coffee shops that are relatively further away from subway stations were still able to garner positive sentiment scores, which begs the question: do location attributes (like proximity to subway stations) really matter as much as coffee attributes (like food and service quality) when it comes to sentiment scores?
How to explore : Hovering over any word will show the number of times the word was used in the text reviews. Correspondingly, the size of the words relates to the number of times it was used; the bigger the word, the more times it was used, and vice versa.
Quick takes : Coffee attributes dominate the reviews. For instance, you see words like delicious, latte, food, friendly, service and milk appearing more. Of course, location attributes still appear, but with reduced frequency—we see outdoor, neighborhood and location—as compared to coffee attributes.